...is proportional to actual wear? (Assuming you think this.). Ignore statistical variations.
Why do you think UOA "wear"...
- Thread starter JAG
- Start date
- Status
I'm tired of hearing UOA wear is porportionate to actual wear.
J/K
I dont think it is.
J/K
I dont think it is.
Vs. chemical etching/dissolving?
Mainly because the metals in solution roughly follow the wear areas + a boost whenever there is contact - Fe for example, or Cu when there is a brass/Cu oil cooler. Except why are Al numbers NOT higher in modular Al engines? Postulation of above makes me believe UOA detected metals come from wear areas. I'm sure the Dr. can find a couple SAE papers on the subject.
Mainly because the metals in solution roughly follow the wear areas + a boost whenever there is contact - Fe for example, or Cu when there is a brass/Cu oil cooler. Except why are Al numbers NOT higher in modular Al engines? Postulation of above makes me believe UOA detected metals come from wear areas. I'm sure the Dr. can find a couple SAE papers on the subject.
- Joined
- Sep 28, 2002
- Messages
- 39,799
Well, if the readable particle distribution has any incremental or decremental cadence to it in relation to total metals...it should be in proportion to mass lost. It may be 100:1 ..but who knows
btw- I have asked Blackstone if they would digest a sample for a reading. They weren't interested. It's not complicated ..but then again, when I saw it done, it was in an in-company environmental lab ...and not a for profit lube testing lab.
Care to chip in if I find anyone??
btw- I have asked Blackstone if they would digest a sample for a reading. They weren't interested. It's not complicated ..but then again, when I saw it done, it was in an in-company environmental lab ...and not a for profit lube testing lab.
Care to chip in if I find anyone??
buster, the first sentence of that quote is true. I'm not sure if the second sentence is.
Gary, what do you mean by digest a sample?
Hello Pablo and Sarge.
I should have got more responses by now because I see people everyday on here concluding that some oil protects actual wear better than another purely because of lower wear numbers in UOAs. They should be able to explain why think they can make that conclusion.
Gary, what do you mean by digest a sample?
Hello Pablo and Sarge.
I should have got more responses by now because I see people everyday on here concluding that some oil protects actual wear better than another purely because of lower wear numbers in UOAs. They should be able to explain why think they can make that conclusion.
- Joined
- Sep 28, 2002
- Messages
- 39,799
Right now we get ppm indications (IIRC) of the 5um and lower "free" metals in the oil. We can't see anything bigger. If you digest the sample (add acid -nitric, I believe), you reduce ALL metals to the particle level ..and they're therefore ALL readable. All the stuff that shows up in PC is all
One guy on JU was a research student and had a really advanced lab (that wasn't an automotive lab) do his UOA. He got metals in the thousands of ppm. Like 2400 Fe and whatnot. It REALLY showed how bypass filters (an Oil Guard in his testing) reduced total metals by about 40% (IIRC) when compared to various full flows.
This is done in environmental testing where you differentiate between "free" Cu and "total" Cu in effluents and whatnot. To describe the difference ...one you could have a penny in the sample and ONLY read what gets suspended in the sample off of the surface ..the other the entire penny would be dissolved and suspended. Radically different view.
I'm seeing if Terry knows of such a lab. I'd like to show the membership the magnitude of the difference between what we see ..and what truly is there. Naturally ...everyone's mileage may vary.
In this type of testing, you would truly see whatever unfiltered mass was contained in the oil. It won't necessarily tell you where it came from ..but should be a closer indicator of actual wear rates.
That's very cool Gary. I'll chip in money for it if enough people do to get the cost to a reasonable amount per person.
As I've said in past threads, in my oil testing, a certain oil (Lubromoly 0W-40) when heated above 300F for a few hours, dissolved a lot of copper off a penny and I could see it like a muddy river emptying into a clear ocean. It was wild.
As I've said in past threads, in my oil testing, a certain oil (Lubromoly 0W-40) when heated above 300F for a few hours, dissolved a lot of copper off a penny and I could see it like a muddy river emptying into a clear ocean. It was wild.
I've seen this point made here before. But I've yet to see an answer supported by controlled testing or a scientific paper.
Some in the industry may have access to enough UOA data to make accurate accessments of an oil's effectiveness at protecting against actual wear in certain engines, but it doesn't exist here. And if there is a scientific paper on the subject, it deserves mention at BITOG, especially considering the emphasis and importance placed on the supposed relationship between UOAs and engine wear in healthy engines that runs rampent on this site.
My two cents here is that given just how penny pinching most are on this site with the cost of the lubricant going into one's crankcase, why on earth would they put forth the funds for the cost of a test that thorough and complete?
That is why although it may seem like I'm wasting a perfectly good load of synthetic lubricant in my engine, to me, it makes more economical sense just to change the oil out a little more often, than to fork out for the UOA, just to find out that the oil is shot. Then I would be out the cost of the analysis AND the new oil for the engine.
Now I'm not saying that in fleet applications or business applications where down time is a big issue, that a test like this really would help the cause and actually lower operating costs by really seeing exactly how far the oil can go in offering great protection, while keeping wear metals in check.
That is why although it may seem like I'm wasting a perfectly good load of synthetic lubricant in my engine, to me, it makes more economical sense just to change the oil out a little more often, than to fork out for the UOA, just to find out that the oil is shot. Then I would be out the cost of the analysis AND the new oil for the engine.
Now I'm not saying that in fleet applications or business applications where down time is a big issue, that a test like this really would help the cause and actually lower operating costs by really seeing exactly how far the oil can go in offering great protection, while keeping wear metals in check.
Brian, thank you for the well written post which mimics my thoughts.
Dad2leia, I'm also a penny pincher but am interested enough in scientific findings of this sort enough to overcome my penny-pinching ways if the cost is reasonable.
We've yet to hear from anyone who can justify what I've asked people to justify. This thread hasn't been open for long but I wanted to mention the thread's purpose again so it doesn't get too off-track.
Dad2leia, I'm also a penny pincher but am interested enough in scientific findings of this sort enough to overcome my penny-pinching ways if the cost is reasonable.
We've yet to hear from anyone who can justify what I've asked people to justify. This thread hasn't been open for long but I wanted to mention the thread's purpose again so it doesn't get too off-track.
It seems like at best it would depend. If you're managing lubricants for a race team and are using fresh engines all the time then what you see in the oil probably represents what is being produced by wear, provided you've done a flush of some sort to clean the engine.
With engines that have been run for awhile it gets more complicated. Some studies relying upon something in the oil use pretty elaborate flushing procedures when changing types of oil, as there is a 'residue' of chemicals and metals in engines.
A 'residue' seems like common sense, considering that we rely upon film formation for wear protection, and that we also end up with 'varnish' and such in engines. In a FAQ on their aircraft oil web site Mobil-Exxon mentions that when switching to their synthetic blend one will see elevated metals for awhile, I guess in part due to existing films being dissolved and/or replaced.
The best data that I've seen on the subject was from one doing UOAs on a regular basis on a long Mobil 1 oil change interval, where they also tracked oil consumption. Most people seem to just average the metals values over entire interval, but averages by design minimize variation and in some cases it's the variation that you're most interested in. As an example long term averages can hide the fact that your oil was toasted three samples back. Looking at the data by metals introduced between each UOA sample one sees a pattern of an initial spike, then it gets low, then it goes up, and eventually the data doesn't make sense as you end up with negative values of metal being introduced. For me this confirmed that something, probably film(s) formation, was soaking up some of the metals, and that UOA accuracy may not be adequate for some types of conclusions that are needed.
In summary, the practice of averaging UOAs over a long period seems so ingrained that it's almost impossible to even explain in this forum why it's a problem, much less show data that demonstrates it. Negative wear metals betwen UOA samples makes it clear that typical UOAs aren't reliable.
With engines that have been run for awhile it gets more complicated. Some studies relying upon something in the oil use pretty elaborate flushing procedures when changing types of oil, as there is a 'residue' of chemicals and metals in engines.
A 'residue' seems like common sense, considering that we rely upon film formation for wear protection, and that we also end up with 'varnish' and such in engines. In a FAQ on their aircraft oil web site Mobil-Exxon mentions that when switching to their synthetic blend one will see elevated metals for awhile, I guess in part due to existing films being dissolved and/or replaced.
The best data that I've seen on the subject was from one doing UOAs on a regular basis on a long Mobil 1 oil change interval, where they also tracked oil consumption. Most people seem to just average the metals values over entire interval, but averages by design minimize variation and in some cases it's the variation that you're most interested in. As an example long term averages can hide the fact that your oil was toasted three samples back. Looking at the data by metals introduced between each UOA sample one sees a pattern of an initial spike, then it gets low, then it goes up, and eventually the data doesn't make sense as you end up with negative values of metal being introduced. For me this confirmed that something, probably film(s) formation, was soaking up some of the metals, and that UOA accuracy may not be adequate for some types of conclusions that are needed.
In summary, the practice of averaging UOAs over a long period seems so ingrained that it's almost impossible to even explain in this forum why it's a problem, much less show data that demonstrates it. Negative wear metals betwen UOA samples makes it clear that typical UOAs aren't reliable.
Good post 1sttruck. It is complicated for sure...many things that can occur that make wear metal levels in UOAs not representative of actual wear. Sometimes overestimating it and sometime underestimating it.
The particle size detection limits of ICP (mentioned in buster's quote) used to measure wear metal levels in UOAs in one of the factors that can cause this misleading I mentioned. I've rubbed off ferrous wear particles captured from my various magnets I use in the engine and with the aid of a microscope measured the particle sizes of the larger particles captured. Many of them were way too big for ICP to detect IF they had made it into the oil sample bottle. In the UOA forum, I have a thread or two showing this.
Wear metal "fallout" onto engine parts is another factor. This will depend on the oil used and its condition as well as the engine and its condition (cleanliness and mechanical condition), as well as the size of wear metals generated. When the oil is drained and the oil sample taken, it does not capture what would be there if the engine had had a perfect scrubbing and mixing just prior to draining the oil. Dissolving and corrosion of metals is another factor which also depends on the oil used, its condition, the engine and its condition. These particles tend to be very small and thus are not only well-suspended in the oil but also accurately measured by ICP. I could go on but that's enough rambling from me for now...back to work.
The particle size detection limits of ICP (mentioned in buster's quote) used to measure wear metal levels in UOAs in one of the factors that can cause this misleading I mentioned. I've rubbed off ferrous wear particles captured from my various magnets I use in the engine and with the aid of a microscope measured the particle sizes of the larger particles captured. Many of them were way too big for ICP to detect IF they had made it into the oil sample bottle. In the UOA forum, I have a thread or two showing this.
Wear metal "fallout" onto engine parts is another factor. This will depend on the oil used and its condition as well as the engine and its condition (cleanliness and mechanical condition), as well as the size of wear metals generated. When the oil is drained and the oil sample taken, it does not capture what would be there if the engine had had a perfect scrubbing and mixing just prior to draining the oil. Dissolving and corrosion of metals is another factor which also depends on the oil used, its condition, the engine and its condition. These particles tend to be very small and thus are not only well-suspended in the oil but also accurately measured by ICP. I could go on but that's enough rambling from me for now...back to work.
JAG,
One idea example of UOA wear metals not representing the total actual wear is from a lawnmowever first fill that I posted a couple months ago. The wear metals were all fairly low.
One idea example of UOA wear metals not representing the total actual wear is from a lawnmowever first fill that I posted a couple months ago. The wear metals were all fairly low.
Yeah, I remember that. Thanks for reminding me. I dug it up and here it is: http://theoildrop.server101.com/forums/s...true#Post964285
Ferrous parts will collect a film of magnetic particles at the boundaries of magnetic domains, which might cause varying Fe readings.
Title: Fire Ring Wear Assessment on a 6v92ta Detroit Diesel Engine
Document Number: 890420
Author(s):
R. A. Simmons - Petro-Canada Products
K. R. Dymock - Petro-Canada Products
G. D. Webster - National Research Council of Canada
Abstract:
Initial work has been completed on a fire ring (top piston ring) wear assessment in a military pattern heavy duty high speed diesel running on a range of fuel qualities............With the fire rings as the only major sources of chromium from engine wear components, engine oil analysis for trace chromium (by atomic absorption) in the 50-1500 ppb range could then be used to determine fire ring wear............. The results were verified by data obtained earlier from 500 hour durability runs in which the fire rings were physically weighed. The wear measurements made by the oil analysis technique correlated within 10% of the 500 hour measurements...............
Document Number: 890420
Author(s):
R. A. Simmons - Petro-Canada Products
K. R. Dymock - Petro-Canada Products
G. D. Webster - National Research Council of Canada
Abstract:
Initial work has been completed on a fire ring (top piston ring) wear assessment in a military pattern heavy duty high speed diesel running on a range of fuel qualities............With the fire rings as the only major sources of chromium from engine wear components, engine oil analysis for trace chromium (by atomic absorption) in the 50-1500 ppb range could then be used to determine fire ring wear............. The results were verified by data obtained earlier from 500 hour durability runs in which the fire rings were physically weighed. The wear measurements made by the oil analysis technique correlated within 10% of the 500 hour measurements...............
Thanks. Chromium on rings are uniquely hard (making wear particles small) and also uniquely resistant to corrosion and chemical dissolution, all of which makes it quite accurately measured in UOAs. Studies similar to the above on the other main wear metals is what's additionally needed.
The Fire Ring Wear Assessment document referred to above appears to be quite different from the approach used here. According to the abstract, the variables were fuel quality, injection type, and injection timing, and it appears that the testing was done under controlled (8 hour) test conditions. UOAs are great for determining the condition of used oil, and can be used to uncover problems in engines. That’s why it’s used during engine development testing to uncover potential durability issues like the ones in the paper referenced above.
But that’s not what is typically happens here. At BITOG, many compare a few UOAs from a healthy engine or engines run under variable conditions (which can include different drivers, fuel, weather, etc.) with different oils and conclude which oil is best based on what is often times very low wear numbers with just a few PPM difference. With that approach, the conclusions are nearly always based on data buried in the “noise” of the “test” results. That makes the conclusions suspect.
But that’s not what is typically happens here. At BITOG, many compare a few UOAs from a healthy engine or engines run under variable conditions (which can include different drivers, fuel, weather, etc.) with different oils and conclude which oil is best based on what is often times very low wear numbers with just a few PPM difference. With that approach, the conclusions are nearly always based on data buried in the “noise” of the “test” results. That makes the conclusions suspect.
- Status
Similar threads
